Because exposures to carcinogens arise from many sources and routes, epidemiology studies should use biomarkers of internal dose as surrogates for carcinogen exposures. However, direct measurement of carcinogens is impractical because these chemicals are usually reactive electrophiles with short life spans in the body. As an alternative, carcinogen doses can be investigated via addition products (adducts) of the electrophiles with blood proteins, especially human serum albumin (HSA) and hemoglobin (Hb). Because these adducts are not repaired, they reflect doses of carcinogens over one to two months. In this project, we envision a 'protein adductome', representing all adducts present on a given protein, as an entity of toxicogenomic importance. Since all electrophilic species are potentially carcinogenic, the protein adductome is arguably more relevant to carcinogen discovery than the proteome or the metabolome. Despite their utility as biomarkers of carcinogen dose, adducts of toxic chemicals with abundant blood proteins, such as human serum albumin (HSA) and Hb, have rarely been used in epidemiology studies or to discover the initiators of human cancers. There are three major reasons for this. First, without substantial adduct enrichment, typical concentrations of HSA and Hb adducts are too low to permit widespread * discovery of new carcinogens. This is the 'needle-in-a-haystack problem'that has also plagued proteomic investigations. Second, state-of-the-art mass spectrometry (MS) has not been exploited to identify unknown protein adducts or to profile adducts with links to human cancers. And third, it is difficult to obtain blood samples with which to assay protein adducts. In this project, we hypothesize that particular subadductomes, such as the free cysteine of HSA (Cys34), can be used to discover new carcinogens and to serve as biomarkers of internal dose in large epidemiology studies. To consider these questions, we will selectively enrich cysteinyl adducts of HSA from human blood and then will use MS to characterize the known and unknown HSA adducts. By comparing adduct maps between lymphoma cases and control subjects, we will pinpoint possible carcinogens. Regarding the use of protein adducts in epidemiology studies, we will collaborate with Project 3 to determine whether it is feasible to quickly measure protein adducts of polycyclic aromatic hydrocarbons (PAH) in a single drop of blood.